You are on page 1of 11

Engineering Geology 79 (2005) 151 161 www.elsevier.


A size classification for debris flows

Matthias JakobT
BGC Engineering Inc., 500-1045 Howe Street, Vancouver, British Columbia, Canada Received 24 February 2004; received in revised form 30 December 2004 Available online 24 February 2005

Abstract A 10-fold classification for debris flow size is proposed based on total volume, peak discharge and area inundated by debris. Size classes can be used for regional overview studies where detailed site investigations are either unnecessary, too costly or where the highest hazard and risk creeks need to be identified for further study. They are also useful to compare the regional impact between affected areas and the effects of rainstorms, and they allow lay-people to obtain an understanding of debris flow magnitude and consequences. Finally, different size classes allow the estimation of travel times to points of interest based on empirically derived equations. It is proposed that agencies concerned with debris flows should establish a documentation of debris flow size according to this classification, which serves as a data base for hazard and risk planning. D 2005 Elsevier B.V. All rights reserved.
Keywords: Debris flow; Lahar; Landslide; Size classification; Hazard; Risk

1. Introduction Existing classifications for landslides are based on process, morphology, geometry, movement type and rate, type of material and activity (Varnes, 1978; Cruden and Varnes, 1996; Hungr et al., 2001). Size classifications are rarely used for landslide characterization because they provide too little information on other, often more important, morphologic or process characteristics of a landslide. This paper argues that for debris flows a classification that incorporates

T Tel.: +1 604 684 5900x112; fax: +1 604 684 5909. E-mail address: 0013-7952/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.enggeo.2005.01.006

different size characteristics could be used in regional studies along infrastructure corridors because it addresses variables that are part of a hazard evaluation. A mandatory debris-flow size classification is proposed as part of provincial and federal landslide inventory initiatives, because these data will yield information where detailed hazard assessments are required. This suggestion is motivated by the common practice of simply removing debris flow deposits by contractors whose scope of work does not include the documentation of debris flow size. The purpose of their work is to reestablish traffic, electricity or oil and gas flow, or to decrease hazard to landowners. The local government or utility provider responsible for the clean-up


M. Jakob / Engineering Geology 79 (2005) 151161 Table 2 Size classification for snow avalanches (McClung and Schaerer, 1981) Size class 1 2 3 Typical mass (Mg) b10 102 103 Path length (m) 10 100 1000 Impact pressures (kPa) 1 10 100 Potential consequences

rarely maintains an inventory of size variables, or event dates (there are some exceptions, for example, for rockfall). Several European countries have also begun to establish a systematic and coordinated documentation of mountain disasters, which will aid future research and cost-effective disaster mitigation (Huebel et al., 2002). The general neglect of recording debris flow size contrasts to routine collection of stream flow and precipitation data by most governments. Collection of these data is often limited to scientists or engineers who volunteer their time to further investigate a debris flow event, or are involved as expert witness if legal action ensues. This occurs some time after an event has taken place and debris flow size data are usually of lesser quality. This paper discusses some size classifications for slope hazards and the reason why a separate classification is warranted for debris flows. The new size classification is introduced by explaining the choice of variables and defining the individual classes. An application of the system is presented for the analysis of a road through debris-flow-prone terrain followed by a discussion of the classification system with regard to size dependency for different return periods.







Relatively harmless to people Could bury, injure or kill a person Could bury a car, destroy a small building or break a few trees Could destroy a large railway car or truck, several buildings or a forest with an area up to four hectars (4000 m2) Largest avalanches known could destroy a large village or a forest of 40 ha

2. Previous work on slope hazard size classifications A landslide size classification based on volume was proposed by Fell (1994) (Table 1). Volume, however, is of minor relevancy when pipeline, railway or highway crossings are considered in a hazard or risk analysis. Furthermore, the size ranges appear to be arbitrary and could be simplified by a classification system based on powers of 10.
Table 1 Size classification for landslides (Fell, 1994) Size class 1 2 3 4 5 6 7 Size description Extremely small Very small Small Medium Mediumlarge Very large Extremely large Volume (m3) b500 5005000 500050,000 50,000250,000 250,0001,000,000 1,000,0005,000,000 N5,000,000

Pierson (1998) included a size classification to differentiate travel times of wet volcanic mass flows by discharge because of a positive correlation between velocity and debris flow magnitude. He differentiated between moderate (102 to 103 m3/s), large (103 to 104 m3/s), very large (104 to 106 m3/s) and extremely large (N106 m3/s) volcanic debris flows. A well-documented size classification system exists for snow avalanches, which has been applied for over 30 years (de Quervain et al., 1973). European approaches focused on relative size classes such as proposed by Fohn (1975). A more quanti tative system was proposed by Fohn et al. (1977) with a three-fold classification of dsmallT, dmediumT and dlargeT based on avalanche runout area. The main criticism is that some important variables are not included and the insufficient number of classes to differentiate between events (McClung and Schaerer, 1981). In 1977, the Canadian Avalanche Committee adopted a modified size classification system originally introduced by the USDA (1961) to estimate the potential destruction in the middle of the avalanche path at terminal velocity. This system identifies hazard potential of differently sized events and thus provides input to scientific studies, hazard and risk analyses or mitigation design. Table 2 lists the factors used in the size classification for avalanches.

M. Jakob / Engineering Geology 79 (2005) 151161


3. Debris-flow size classification The basic requirement of a size classification of debris flows is that at least some variables included are easily obtainable, are meaningful for assessing hazard and risk, and can be used for evaluating the need for passive or active debris flow mitigation. As demonstrated by Pierson (1998), they are also useful in predicting travel times for debris flows and may thus allow timely evacuation or closure of transportation corridors. The variables proposed in this paper include total debris volume, peak discharge and area inundated, and a description of potential consequences (Fig. 1). While measures of size may be difficult to conceive for the lay-person, descriptions of potential impact are easily understood. The system proposed in this paper does not include descriptive size classes such as bsmallQ, bmediumQ and blargeQ as these descriptors are very vague, and can only obtain objectivity if used in connection with numbers. The adoption of other size classifications, for example the existing one for snow avalanches, is not appropriate due to significant differences in their use. These differences pertain to the physical properties of the materials, the number of classes, the types of meaningful variables chosen and the need to

differentiate between volcanic and other muddy debris flows and bouldery debris flows. Debris flows also span a much wider range of volumes due to the much greater volume of available materials, especially from Quaternary volcanoes. Volcanic debris flows from Mount Rainier, for example, have reached volumes of up to 3109 m3 (Scott et al., 1995). This much wider size range justifies the introduction of additional classes. 3.1. Parameters used in the size classification The use of debris flow volume (V) only in describing the size of a debris flow may be insufficient to allow its use for hazard and risk assessments. For these studies it is important to determine the peak discharge ( Q) and the area that will likely be inundated by debris (B). Debris volume is a variable used for the design of debris basins or debris barriers whose purposes are to halt the debris flow before reaching an area of high potential consequences. Debris volume also affects runout distance and area covered by a debris flow and is needed as an input parameter in existing runout models (OBrien and Julien, 1989; Hungr, 1995; Iverson et al., 1998; Iverson and Denlinger, 2001).

Fig. 1. Conceptual sketch showing the variables on which the debris-flow size classification is based.


M. Jakob / Engineering Geology 79 (2005) 151161 Table 3 Correlations between peak discharge ( Q p) And total volume (V) Formula Q p=0.135V 0.78 (bouldery debris flows) Q p=0.019V 0.79 (muddy debris flows) Q p=0.006V 0.83 (volcanic debris flows) Q p=0.04V 0.90 (bouldery debris flows) Q p=0.003V 1.01 (volcanic debris flows) Q p=0.001V 0.87 (volcanic debris flows) Q p=0.1V 0.83 (bouldery debris flows) Author Mizuyama et al. (1992) Mizuyama et al. (1992) Jitousono et al. (1996) Bovis and Jakob (1999) Bovis and Jakob (1999) Jitousono et al. (1996) Rickenmann (1999)

Therefore, volume directly affects debris flow hazard or intensity maps and ultimately derivative land-use maps, which makes it a more appropriate measure than mass. Debris-flow volume is defined as the total volume of debris transported beyond a point of interest (usually the apex of a creek fan on which development occurs). It should be noted, however, that deposit thickness is often very difficult to determine. Peak discharge is an input variable for the design of bridges, culverts, pipeline crossings and channelization works as it determines the cross-section area needed to pass the design flow. Discharge is also needed as an input parameter for models that require a hydrograph as an input variable (OBrien, 2001). Peak discharge is defined as the product of the largest cross-section in a reach of interest and the highest estimated or measured flow velocity of the flow passing this cross-section. Flow cross-section can easily be measured, while velocity estimates can be achieved by a number of methods discussed by Jakob (2005). Peak discharge has been correlated with total debris volume for different types of debris flows (Costa, 1988; Mizuyama et al., 1992; Bovis and Jakob, 1999; Jitousono et al., 1996; Rickenmann, 1999). The formulae developed by these authors provide a range of possible peak discharges for muddy or volcanic and bouldery debris flows associated with the respective volume. The wide range of debris flow discharge can at least be partially explained by the frictional resistance of coarse boulder fronts, which can slow debris flows to velocities significantly less than without those boulder fronts. The highest and lowest discharge values from the equations in Table 3 were used to define the discharge classes in Table 4. Using the size classification summarized in Table 4, debris flow volume can then be estimated from peak discharge or peak discharge can be determined from volume. Finally, area inundated is included because it provides a measure of debris flow mobility and potential consequences. For example, a developer requires inundation area information for the zonation of a resort in a valley or on a fan, while an engineering company requires B to design efficient deflection berms for potential debris flows. Area inundated was again separated in bouldery debris flows and muddy (mostly volcanic) debris flows because the latter will

Caution should be exercised for the differentiation between granular debris flows and muddy debris flows for size classes exceeding 10,000 m3 in volume (Class 4 and 5). Work by Rickenmann (1999) has demonstrated that regression lines for some muddy flows and granular flows can converge at these higher classes. Locally derived relationships between volume, peak discharge and area inundated may be used for an improved assignment of debris flow size into the correct classes.

spread over larger areas due to their higher mobility. Areas inundated, B v, by volcanic debris flows were calculated by using the relationship: Bv 200V 2=3 (Iverson et al., 1998) where V equals the debris flow volume, and: Bb 20V 2=3 (Griswold, 2004) 2 1

for bouldery debris flows. Path length, which is defined as the total travel distance from the initiation of the debris flow to its farthest runout, is not used due to its poor correlation with debris-flow volume. Similarly, runout distance, defined as the total distance of debris flow deposition, is not included in the size classification because many debris-flow channels end in higher order streams, which truncate the debris flow thus providing an underestimation of runout. In addition, debris flow runout is often assessed very subjectively because some workers define runout as the limit of fine sediment deposition, while others use the extent of coarse debris. 3.2. Size definitions This section describes and justifies the size classes that have been used in this classification. It is necessary to point out that volume estimates reflect the total volume transported beyond a point of interest (usually the fan apex). The peak discharge is considered near the fan apex and the area inundated

M. Jakob / Engineering Geology 79 (2005) 151161 Table 4 Size classification for debris flows Size class 1 2 3 4 5 6 7 8 9 10 V, range (m3) b102 102103 103104 104105 105106 105106 106107 107108 108109 N109 Q b, range (m3/s) b5 530 30200 2001500 150012,000 N/A N/A N/A N/A N/A Q v, range (m3/s) b1 13 330 30300 3003103 31033104 31043105 31053106 31063107 31073108 Bb (m2) b4102 41022103 21039103 91034104 41042105 N2105 N/A N/A N/A N/A Bv (m2) b4103 41032104 21049104 91044105 41052106 21063107 31073108 31083109 310931010 N31010 Potential consequences


Very localized damage, known to have killed forestry workers in small gullies, damage small buildings Could bury cars, destroy a small wooden building, break trees, block culverts, derail trains Could destroy larger buildings, damage concrete bridge piers, block or damage highways and pipelines Could destroy parts of villages, destroy sections of infrastructure corridors, bridges, could block creeks Could destroy parts of towns, destroy forests of 2 km2 in area, block creeks and small rivers Could destroy towns, obliterate valleys or fans up to several tens of km2 in size, dam rivers Could destroy parts of cities, obliterate valleys or fans up to several tens of km2 in size, dam large rivers Could destroy cities, inundate large valleys up to 100 km2 in size, dam large rivers Vast and complete destruction over hundreds of km2 Vast and complete destruction over hundreds of km2

V is the total volume, Q b and Q v are the peak discharge for bouldery and volcanic debris flows, respectively, B b and B v are the area inundated by bouldery and volcanic debris flows. N/A signifies that bouldery debris flows of this magnitude have not been observed. The constant in Eq. (2) was rounded so that B by non-volcanic debris flows is 10 times smaller than that of volcanic debris flows.

refers to the debris flow fan. For Class 6 to 10 volcanic debris flows where entire floodplain or mountain forelands are inundated, the three size categories refer to the maximum volume transported beyond the initiation area, the peak discharge anywhere along the transport zone, and the total area inundated downstream of the initiation area. The description of potential consequences is based on an extensive review of the debris flow literature, specifically the proceedings of the Conference on Debris Flow Hazards Mitigation: Mechanics, Prediction and Assessment (Chen, 1997; Wieczorek and Naeser, 2000; Rickenmann and Chen, 2003). Class 1 debris flows (10 to 102 m3 volume) occur in small channels or gullies and usually entrain little material as they flow downhill. They may be triggered in-channel or by a small debris slide or perhaps rock fall. Sediment supply limitations, insufficient water or rapid drainage in coarse and highly permeable colluvial deposits and low gradient encourage early deposition. These types of debris flows can be found in small gullies in hilly or mountainous terrain (Fig. 2). Debris flows with up to 100 m3 volume, discharge of up to 5 m3/s and up to 4000 m2 of inundated area (Table 4), can cause significant damage if structures

are located near the fan apex where the debris flow is still traveling fast and carrying large boulders. There is no class smaller than Class 1 because debris flows with volumes under 10 m3 are usually of limited practical interest. Class 2 debris flows (102 to 103 m3) occur on small creeks that are often, but not necessarily, supplylimited. Debris flows in supply limited basins require a significant recharge period prior to each debris flow event and exhibit a lower frequency of debris flow activity. In contrast, supply-unlimited basins are controlled primarily by hydroclimatic events, since the supply of easily mobilized sediment is rarely a limiting condition for debris flow occurrence (Fig. 2). With a peak discharge of up to 30 m3/s, Class 2 debris flows can destroy small brick or wooden buildings over an area of up to 20,000 m2 in the runout zone. Very heavy rainfall events can cause hundreds or thousands of Class 2 and Class 3 debris flows in a large region that may coalesce into higher class events (e.g. Lopez et al., 2003; Jan and Chen, 2005). Class 1 and 2 are rarely described in the scientific literature because case studies focus on spectacular or destructive events with few exceptions (Johnson and Warburton, 2003).


M. Jakob / Engineering Geology 79 (2005) 151161

Fig. 2. Conceptual sketch for debris flows of size Class 1 to 3.

Class 3 (V=103 to 104 m3) encompasses a very typical size range that produces debris flows at decadal return periods in non-volcanic, supply-limited watersheds or at higher return periods for supplyunlimited channels (Fig. 2). These flows have high destructive potential with peak discharges of up to 200 m3/s and 90,000 m2 area inundated. These flows are not easily recognized as a hazard because their return period in small (b5 km2) watersheds may span decades and dsilent witnessesT such as boulder lobes, impact scars on trees, or trim lines along channels of past debris flow activity are eroded or overgrow (Hungr et al., 1984). Examples of Class 3 debris flows include the 5000 m3 San Rocco debris flow and the 7000 m3 Bertogna debris flows in Italy (Bianco and Franzi, 2003). Class 4 debris flows (104 to 105 m3) may occur on the same creeks as on Class 3 creeks but at return periods of hundreds of years for supply-limited watersheds (Jakob and Bovis, 1996; Bovis and Jakob, 1999), and they are seldom recognized as a significant hazard by individuals other than debris flow experts (Fig. 3). For many supply-limited watersheds of up to 5 km2 size, the upper limit of Class 4 is the maximum size that can be achieved even during very long (thousands of years) return periods. An example of an

upper Class 4 bouldery debris flow with a total volume of 92,000 m3 and a peak discharge of approximately 1000 m3/s is the Hummingbird Creek debris flow in southern British Columbia (Jakob et al., 2000). Class 5 debris flows (105 to 106 m3) originate mostly from volcanoes (but do not need to be syneruptive) or from areas with abundant sediment sources such as the Jiangjia ravine in southern China (Cui et al., 2005). Volcanic debris flows with a high clay content and a poorly developed or non-existent bouldery front that increases flow resistance are significantly more mobile than bouldery debris flows with a sandy-matrix. These characteristics of volcanic debris flows result in excessive runout distance that can convey the hazard from the source area to distant developed areas (Fig. 3). An example of a Class 5 event is the Drift River debris flow with a peak discharge of 2500 m3/s 19 km downstream of the source at Redoubt Volcano that occurred on March 9, 1990 (Dorava and Meyer, 1994). Bouldery debris flows of this magnitude are very rare, and physical evidence of older events can often only be interpreted by analyzing sedimentological evidence in trenches or natural exposures in the runout area. Because of their high volume, discharge of up to 12,000 m3/s and

M. Jakob / Engineering Geology 79 (2005) 151161


Fig. 3. Conceptual sketch for debris flows of size Class 4 to 6.

inundation of up to 2 km2, Class 5 debris flows pose a tremendous risk to infrastructure and populations in their paths (Fig. 3). Class 6 debris flows range from 106 m3 to 107 m3 in volume with a peak discharge of up to 30,000 m3/s and can inundate an area of up to 30 km2. Class 6 debris-flows are usually triggered at volcanoes by edifice flank or sector collapses, eruption-triggered snowmelt or subglacial melting or lake outbreaks (Fig. 3). Class 6 debris flows are responsible for several tens of thousands of lost lives in the past 100 years worldwide (e.g. Wolf, 1878; Murai, 1960; Plafker and Ericksen, 1978; Pierson et al., 1990; Mothes, 1992). Class 6 is also the largest size classification given for bouldery debris flows as larger flows are known only from volcanoes. Four additional classes are introduced for the larger volcanic events that differ from the classes applied by Pierson (1998). The 1998 Casita debris flow in Nicaragua serves as an example for a Class 7 debris flow with a volume of 1.8106 m3 (Scott, 2000). Well-studied examples of Class 8 events are the 1980 South Fork Toutle River debris flow with a volume of 1.3107 m3 and a peak discharge of 68,000 m3/s at Mt. St. Helens (Fairchild, 1985), and the debris flow at Huascaran, Peru (V=1.3107 m3) in 1962, and again in 1970 (V=5

to 10107 m3) (Plafker and Ericksen, 1978). Class 9 debris flows include the 4000 years BP event at Mt. Meager, British Columbia with an estimated volume of 12108 m3 (Friele and Clague, 2004) and the North Fork Toutle River debris flow (V=1.4108 m3, Q=7200 m3/s) at Mt. St. Helens (Fairchild, 1985). Examples of Class 10 debris flows include the Osceola Mudflow with 34 km3 volume and the Electron Mudflow (0.25 km3 volume) at Mount Rainier (Vallance and Scott, 1997). Without sufficient warning, debris flows of Class 7 to 10 could cause thousands to hundreds of thousands of fatalities and regional-scale destruction with long-term socio-economic consequences (Scott et al., 2001). 3.3. Application This section addresses how the size classification could be used for a hazard assessment along a new road, railway or pipeline through debris flow-prone terrain. The first step in such study would likely be a detailed air photograph interpretation along the highway alignment to identify debris-flow-prone creeks. Depending on the photo scale used, photogrammetry can aid in determining channel length (including


M. Jakob / Engineering Geology 79 (2005) 151161

debris-flow-prone tributaries), and channel width. Spot checks in the field allow a rough estimate of yield rate (amount of material stored per meter channel length), and average depth of shallow debris-flow triggering landslides. Area measurements of landslide scars that triggered debris flows can also be accomplished using photogrammetric methods. In many areas and different climate zones the depth of shallow landslides does not differ significantly, which implies that volumetric estimates of the initiating failure can be attempted. Yield rate can be extrapolated over the channel length and summed with expected point source volumes to yield an estimate of debris flow volume. Using this method it is not possible to assign specific return periods for the volume thus determined, which would only be possible by detailed field investigations. This point is further discussed in Section 4. However, the application of the same methods to all watersheds

provides a comparable standard without the necessity of considering return periods. Recognizing that there will always be measurement error, a size classification based on order-of-magnitude estimates will smooth out any large discrepancies that may occur between practitioners. Having estimated a total debris flow volume, Table 4 provides information on the expected range of peak discharge and total area inundated. Air photograph interpretation can usually identify the deposition zone of debris flows (typically downstream of the fan apex), and an approximation can be made as to the location of the area that would likely be debris covered. These data can then be used to rank debris-flow-prone watersheds by their hazard potential, to realign the road to avoid the highest hazard areas and to determine the magnitude of expenditures for debris flow mitigation measures where the hazard can not be avoided.


Class 5 debris-flows (debris avalanche-generated) 1,000

hyperconcentrated flow and normal flood flow

100,000 10000 1000 100 10

Annual Exceedence Probability (years)

Fig. 4. Size classification for debris flows of a single stream (Cheekye River). Particularly for volcanic debris flows a wide spectrum of sizes can be observed, while supply-limited watersheds are likely to produce debris flows in only one or two size ranges.

Debris Flow Peak Discharge (m3/s)

Class 6 debris-flows (rock avalanche-generated)


Total Debris Flow Volume (m3)

M. Jakob / Engineering Geology 79 (2005) 151161


For existing roads, data on past debris flows should be included in the size class estimates. The local government authority responsible for clean-up or restoration after debris flows should request the collection of debris flow size data immediately after an event has occurred. These data could be stored in a web-based database. The classification as proposed in this paper could aid this process because it does not necessarily require specialists to estimate sizes. This is even more important since, in most cases, the reopening of a transportation route is of highest priority, and experts are only sought if damage to people or essential infrastructure has occurred. By the time the expert arrives at the site, however, much of the physical evidence has been removed by the cleanup operations.

5. Conclusions A 10-fold classification system for debris flows is proposed based on debris-flow volume, peak discharge and area inundated by debris. A distinction is made between less mobile bouldery debris flows and highly mobile volcanic and other muddy debris flows. This system will be useful for regional debris-flow studies such as those conducted to evaluate the effects of a landslide-triggering storm or for hazard overview studies along infrastructure corridors. It is suggested that this classification be used routinely whenever a debris-flow has been recorded and that size information be entered in a publicly accessible database. These data would be very useful in detailed hazard assessments that require the construction of frequencymagnitude relationships of debris-flows, and it would help to prioritize mitigation work along highways, railways or pipeline corridors to protect infrastructure from damage. Acknowledgements Tom Pierson, Dieter Rickenmann, Kris Holm, Mike Bovis, Oldrich Hungr, Mike Porter, Kevin Scott, and Jim Vallance provided helpful comments on earlier drafts of this paper. Claire Bonin prepared Figs. 2 and 3. References
Bianco, G., Franzi, L., 2003. Comparison of debris-flow events in three contiguous basins, triggered during the same storm event. In: Rickenmann, D., Chen, C.-l. (Eds.), Proceedings of the Third International Conference on Debris Flow Hazard Mitigation: Mechanics, Prediction, and Assessment, Davos. Millpress, Rotterdam, pp. 943 954. Bovis, M.J., Jakob, M., 1999. The role of debris supply to determine debris flow activity in southwestern B. C. Earth Surface Processes and Landforms 24, 1039 1054. Chen, C.L. (Ed.), 1997. First International Conference on Debris Flow Hazards Mitigation: Mechanics, Prediction, and Assessment, San Francisco. American Society of Civil Engineers, New York. Costa, J.E., 1988. Floods from dam failures. In: Baker, V.R., Patton, P.C. (Eds.), Flood Geomorphology. John Wiley and Sons, New York, pp. 439 463. Cruden, D.M., Varnes, D.J., 1996. Landslide types and processes. In: Turner, A.K., Schuster, R.L. (Eds.), Special Report 247 Landslides: Investigation and Mitigation. Transportation

4. Discussion Debris flows of different size classes may occur in the same watershed, usually increasing in size with longer return periods. In particular, channels draining large Quaternary volcanic complexes can produce debris-flows over a wide class range. For example, the Mount Garibaldi volcanic complex near the southern B.C. town of Squamish produces Class 4 and 5 debris flows on a decadal time scale and Class 6 debris flows on a millenium time scale (Fig. 4) (Friele and Clague, in press). Similarly, at Mount Rainier, debris flows of Class 5 occur almost annually, Class 6 on a century time scale, and Class 7 to 9 on a millenium time scale (Vallance, pers. comm., 2003). The classification of debris-flow size, therefore, should not be interpreted as indicating that a certain class (that of the last reported event) cannot be exceeded. This can only be resolved if a frequencymagnitude relationship is constructed from a detailed reconstruction of past events. Such analysis needed for hazard quantification and delineation may entail separate analyses for debris flows of similar origin. Fig. 4 shows a frequency magnitude curve for Cheekye River, which likely includes debris flows from volcanic rock avalanches that transform into debris flows (Class 6), and smaller debris flows generated by transformation of small debris slides or debris avalanches (Class 4 and 5) (Friele and Clague, in press).


M. Jakob / Engineering Geology 79 (2005) 151161 Jakob, M., Bovis, M.J., 1996. Morphometrical and geotechnical controls of debris flow activity, southern Coast Mountains, British Columbia, Canada. Zeitschrift fqr Geomorphologie Supplementband 104, 13 26. Jakob, M.J., Hungr, O., Anderson, D., Fuller, T., Ayotte, D., 2000. The July 7, 1997 debris flow at Hummingbird Creek. Canadian Geotechnical Journal 37, 1109 1125. Jan, C.D., Chen, C.L., 2005. Debris flows caused by Typhoon Herb in Taiwan. In: Jakob, M., Hungr, O. (Eds.), Debris-Flow Hazard Assessments and Related Phenomena. Chapter 21. Springer Verlag, Heidelberg. in print. Jitousono, T., Shimokawa, E., Tsuchiya, S., 1996. Debris flow following the 1994 eruption with pyroclastic flows in Merapi volcano, Indonesia. Journal of Japan Society Erosion Control Engineering 48 (special issue), 109 116. Johnson, R.M., Warburton, J., 2003. Regional assessment of contemporary debris-flow activity in Lake District mountain catchments, northern England: occurrence, scale and process. In: Rickenmann, D., Chen, C.-l. (Eds.), Proceedings of the Third International Conference on Debris Flow Hazard Mitigation: Mechanics, Prediction, and Assessment, Davos. Millpress, Rotterdam, pp. 965 976. Lopez, J.L., Perez, D., Garcia, R., 2003. Hydrologic and geomorphologic evaluation of the 1999 debris-flow event in Venezuela. In: Rickenmann, D., Chen, C.-l. (Eds.), Proceedings of the Third International Conference on Debris Flow Hazard Mitigation: Mechanics, Prediction, and Assessment, Davos. Millpress, Rotterdam, pp. 989 1000. McClung, D.M., Schaerer, P.A., 1981. Snow avalanche size classification. Proceedings of Avalanche Workshop 35 November, 1980, Vancouver, B.C., Technical Memorandum, vol. 133. The Mountaineers, Seattle, Washington. Mizuyama, T., Kobashi, S., Ou, G., 1992. Prediction of debris flow peak discharge. Interpraevent. International symposium, Bern, Switzerland, Tagespublikation, vol. 4, pp. 99 108. Mothes, P.H., 1992. Lahars of Cotopaxi Volcano, Ecuador: hazard and risk evaluation. In: McCall, G.J.H., Laming, D.J.C., Scott, S.C. (Eds.), Geohazards, Natural and Man-made. Chapman & Hall, London, pp. 53 63. Murai, I, 1960. On the mudflow of the 1926 eruption of Volcano Tokachi-dake, central Hokkaido, Japan. Bulletin of the Earthquake Research Institute (Japan) 38, 55 70. OBrien, J.S., 2001. FLO-2D Users Manual. Nutrioso Arizona. 83 pp. and appendices. OBrien, J.S., Julien, P.Y., 1989. Laboratory analysis of mudflow properties. Journal of Hydraulic Engineering 114, 877 887. Pierson, T.C., 1998. An empirical method for estimating travel times for wet volcanic mass flows. Bulletin of Volcanology 60, 98 109. Pierson, T.C., Janda, R.J., Thouret, J.C., Borrero, C.A., 1990. Perturbation and melting of snow and ice by the 13 November 1985 eruption of Nevado del Ruiz, Columbia, and consequent mobilization, flow and deposition of lahars. Journal of Volcanological and Geothermal Research 21, 1511 1524. Plafker, G., Ericksen, G.E., 1978. Nevados Huascaran avalanches, Peru. In: Voight, B. (Ed.), Rockslides and Avalanches, vol. 1. Elsevier, Amsterdam, pp. 277 314.

Research Board, National Research Council, Washington, D.C., pp. 36 75. Cui, P., Chen, X., Wang, Y., Hu, K., Li, Y., 2005. Jiangjia Ravine debris flows in southern China. In: Jakob, M., Hungr, O. (Eds.), Debris-Flow Hazard Assessments and Related Phenomena. Chapter 22. Springer Verlag, Heidelberg. in print. de Quervain, M., de Crecy, L., LaChapelle, E.R., Losev, K., Shoda, M., 1973. Avalanche classification. Hydrological Sciences Bulletin XVIII (4), 391 402. Dorava, J.M., Meyer, D.F., 1994. Hydrologic hazards in the lower Drift River basin associated with the 19891990 eruptions of Redoubt Volcano, Alaska. Journal of Geothermal Research 62, 387 407. Fairchild, L.H., 1985. Lahars at Mount St. Helens, Washington. PhD dissertation, University of Washington, Seattle, 374 pp. Fell, R., 1994. Landslide risk assessment and acceptable risk. Canadian Geotechnical Journal 31, 261 272. Fohn, P., 1975. Statistische Aspekte bei Lawinenereignissen. Interpraevent 1975, pp. 293 304. Fohn, P., Good, W., Bois, P., Obled, C., 1977. Evaluation and comparison of statistical and conventional methods of forecasting avalanche hazard. Journal of Glaciology 19 (81), 375 387. Friele, P.A., Clague, J.J., 2004. Large Holocene landslides from Pylon Peak, southwestern British Columbia. Canadian Journal of Earth Sciences 41, 165 182. Friele, P.A., Clague, J.J., in press. Multifaceted hazard assessment of Cheekye Fan, a large debris flow fan in southwestern British Columbia. In: Jakob, M., Hungr, O. (Eds.), Debris Flows and Related Phenomena. Springer-Praxis, Heidelberg. Griswold, J.P., 2004. Mobility Statistics and Hazard Mapping for Non-Volcanic Debris Flows and Rock Avalanches. Unpublished master thesis. Portland State University. Huebel, J., Kienholz, H., Loipersberger, A. 2002. DOMODIS Documentation of Mountain Disasters: State of the discussion in the European Mountain Areas, Handbuch 1. Schriftenreihe 1; Internationale Forschungs-gesellschaft INTERPRAEVENT, Klagenfurt, Austria. Hungr, O., 1995. A model for the runout analysis of rapid flow slides, debris flows, and avalanches. Canadian Geotechnical Journal 32, 610 623. Hungr, O., Morgan, G.C., Kellerhals, R., 1984. Quantitative analysis of debris torrent hazards for design of remedial measures. Canadian Geotechnical Journal 21, 663 677. Hungr, O., Evans, S.G., Bovis, M.J., Hutchinson, J.N., 2001. A review of the classification of landslides in the flow type. Environmental and Engineering Geoscience VII (3), 221 228. Iverson, R.M., Denlinger, R.P., 2001. Flow of variably fluidized granular masses across three-dimensional terrain: 1. Coulomb mixture theory. Journal of Geophysical Research 106 (B1), 537 552. Iverson, R.M., Schilling, S.P., Vallance, J.W., 1998. Objective delineation of lahar inundation zones. GSA Bulletin 110 (8), 972 984. Jakob, M., 2005. Debris flow hazard assessments. In: Jakob, M., Hungr, O. (Eds.), Debris-Flow Hazard Assessments and Related Phenomena. Chapter 17. Springer Verlag, Heidelberg. in print.

M. Jakob / Engineering Geology 79 (2005) 151161 Rickenmann, D., 1999. Empirical relationships for debris flows. Natural Hazards 19, 47 77. Rickenmann, D., Chen, C.L. (Eds.), 2003. Proceedings of the Third International Conference on Debris Flow Hazard Mitigation: Mechanics, Prediction, and Assessment, Davos. Millress, Rotterdam. Scott, K.M., 2000. Precipitation-triggered debris flow at Casita Volcano, Nicaragua: implications for mitigation strategies in volcanic and tectonically active steeplands. In: Wieczorek, G.F., Naeser, N.D. (Eds.), Proceedings of the Second International Conference on Debris Flow Hazards Mitigation: Mechanics, Prediction and Assessment, Taipei. Balkeema, Rotterdam, pp. 3 13. Scott, K.M., Vallance, J.W., Pringle, P.T., 1995. Sedimentology, behaviour and hazards of debris flows at Mount Rainier, Washington. United States Geological Survey Professional Paper 1547 (56 pp.). Scott, K.M., Macas, J.L., Naranjo, J.A., Rodrguez, S., McGeehin, J.P., 2001. Catastrophic debris flows from landslides in volcanic terrains, mobility, hazard assessment, and mitigation strategies. USGS Geological Survey Professional Paper 1630 (59 pp.).


USDA Forest Service, 1961. Snow Avalanches. USDA Handbook No. 194 with revisions in 1968. Washington, D.C. [84 pp.]. Vallance, J.W., 2003. Written communications, December 2. Vallance, J.W., Scott, K.M., 1997. The Osceola mudflow from Mount Rainier: sedimentology and hazard implications of a huge clay-rich debris flow. Geological Society of America Bulletin 109, 143 163. Varnes, D.J., 1978. Slope movement types and processes. In: Schuster, R.L., Krizek, R.J. (Eds.), Special Report 176 Landslides: Analysis and Control. Transportation Research Board, National Research Council, Washington D.C., pp. 11 33. Wieczorek, G.F., Naeser, N.D. (Eds.), 2000. Proceedings of the Second International Conference on Debris Flow Hazards Mitigation: Mechanics, Prediction and Assessment, Taipei. Balkeema, Rotterdam. Wolf, T., 1878. Geognostische Mitteilungen aus Ecuador; Der Cotopaxi und seine letzte Eruption am 26. Juni 1877. Neues Jahrbuch der Mineralogie, Geologie und Pal7ontologie, 113 167.